Tiiberculous Pleural Effusions (n = 5): Diagnosis was assessed in two cases by histologic examination of pleural biopsy specimens and confirmed by culture (d,e), in two others by culture of pleural fluid (b,c). In the last case, pleural fluid stained with auramin showed acid-fast bacilli and the diagnosis was confirmed by culture (this patient also had pulmonary cavitary lesionsXa).
Nontuberculous Pleural Effusions (n = 14): Three patients presented with a pleural transudation associated with a cardiac failure, nine with a neoplastic pleural involvement, one with a pleural reaction secondary to a pulmonary embolism, and one to a pneumonia. Mycobacterial cultures were performed in all cases and remained negative.
Protein concentrations were determined using a protein assay kit and IgG and IgA concentrations were evaluated by immunonephe- lometry.
Preparation of P32: This purified 32 lalodalton protein antigen was prepared from zinc deficient Sauton culture filtrate of Mycobacterium bovis substrain 1173P2 as described previously.
Dot Blot assay: We performed this assay for the serodiagnosis of tuberculous disease. It gives the same results as the better known ELISA. A reflectance densitometer is used to quantify the peroxidase stain on the immunoblots. The titer of each sample was defined as the reciprocal of the highest dilution giving a reflectance measurement higher than a predetermined limit (0.2 reflectance unit).